• Supervised learning (labeling things)

• 2010 - 2020: Large scale supervised learning

• LLM

  • How? supervised learning (A->B) 重複預測下一個word

• 例子

  • 寫作：rewrite for clarity
  • 閱讀：有沒有在抱怨、情緒分析
  • 聊天：聊天機器人

• web search or LLM？

• LLM可能會錯，但回答比較精簡

  • web search有時會得到比較好的答案，但要花時間找到你要的資訊

• setup

import openai
import os

openai.api_key = os.getenv("OPENAI_API_KEY")

def llm_response(prompt):
    response = openai.ChatCompletion.create(
        model='gpt-3.5-turbo',
        messages=[{'role':'user','content':prompt}],
        temperature=0
    )
    return response.choices[0].message['content']
python

• classify

prompt = '''
    Classify the following review 
    as having either a positive or
    negative sentiment:

    The banana pudding was really tasty!
'''

response = llm_response(prompt)
print(response)
python

all_reviews = [
    'The mochi is excellent!',
    'Best soup dumplings I have ever eaten.',
    'Not worth the 3 month wait for a reservation.',
    'The colorful tablecloths made me smile!',
    'The pasta was cold.'
]

all_reviews
classifications = []
for review in all_reviews:
    prompt = f'''
        Classify the following review 
        as having either a positive or
        negative sentiment. State your answer
        as a single word, either "positive" or
        "negative":

        {review}
        '''
    response = llm_response(prompt)
    classifications.append(response)

classifications
python

• w3 meterials
  • he economic potential of generative AI: The next productivity frontier
  • GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models
  • The Potentially Large Effects of Artificial Intelligence on Economic Growth (Briggs/Kodnani)
  • https://www.onetonline.org/

• https://www.coursera.org/learn/generative-ai-for-everyone
• slides: https://community.deeplearning.ai/t/generative-ai-for-everyone-lecture-notes/481740